ൢ֙ـं (ᖪ 24: 293-303 (2004ٿέ៉ Formosan Entomol. 24: 293-303 (2004) The Mosquito Fauna (Culicidae: Diptera) in the Surface Water of Hapen Creek, Ilan, Taiwan Hwa-Jen Teng, Meng-Hau Hsue and Yen-Li Wu Medical Entomological Laboratory, Research and Development Center, Center for Disease Control, Department of Health, Taipei, Taiwan 115, R.O.C. How-Jing Lee* Department of Entomology, National Taiwan University, Taipei, Taiwan 106, R.O.C. ABSTRACT A 1-year survey was conducted to understand the mosquito fauna and seasonal abundance of the dominant species in Hapen Creek in the Fushen Botanical Garden, northern Taiwan. Anopheles bengalensis Puri was found to be the most abundant species in terms of the total number collected and its wide distribution at all sampling sites. Culex hayashii Yamada was next and was collected at three out of four sites. Five mosquito groups at Hapen Creek were recognized. Anopheles bengalensis and Cx. hayashii comprised one group, which was dominant in this survey. The average number of An. bengalensis larvae collected per dip increased from April to a peak in June, and then decreased sharply in July. Another peak was found in November and December. For C. hayashii, the first peak was found in October and another in July. The remaining four groups consisted of Cx. nigropunctatus with Cx. halifaxii; Cx. memeticus with An. gigas baileyi; An. lindesayi lindesayi; and Uranotaenia macferlanei. Anopheles lindesayi lindesayi larvae were collected only from February to June in this survey. Water temperature was significantly associated with the number of C. hayashii larvae collected, while total precipitation was associated with the numbers of An. lindesayi lindesayi and An. gigas baileyi collected. From the species correlation analysis of the community to mosquitoes, hemipteran bugs, copepods, and shells were significantly associated with four species of mosquitoes, while shrimp, naucorids, haliplids, dipteran larvae, and water striders were significantly correlated with three species of mosquitoes. Damselflies and stoneflies were positively correlated with two species of mosquitoes. These positive associations with mosquitoes may imply that the species are important natural enemies of mosquitoes or species which coexist with them. Key words: Anopheles bengalensis, Culex hayashii, stream, seasonal fluctuation *Correspondence address e-mail: [email protected] ! Mosquito Fauna of Hapen Creek 293 Introduction places and some microhabitats of mosquitoes became temporary pools or The mosquito fauna plays a major dried out. So, a variety of microhabitats role in the decomposition process of are found in this creek. The objectives of organic debris and serves as food for our study were to understand the mos- consumers of higher trophic levels in quito fauna and the seasonal abundance aquatic ecosystems. Predation by of the dominant species in Hapen Creek. mosquito larvae affects the population The environmental impacts of some dynamics and community structure of available factors such as water velocities, freshwater planktonic and bacterial natural enemies, and precipitation, on communities (Carpenter and Kitchell, this community were analyzed and are 1984; Carpenter et al., 1985; Elser and discussed. Carpenter, 1988; Cochran-Stafira and von Ende, 1998). Many mosquitoes breed in Materials and Methods flowing streams, which vary in water velocity, water depth, the coverage of Sampling Sites. The headwaters of plant species and floating plants, water- Hapen Creek are located in the Fushan bed types, and water quality (Rejman- Botanical Garden at 400~1400 m in kova et al., 1992, 1993, 1998; Almiron, elevation. The area of this garden is 1996). These factors determine the 1097.9 ha which it is divided into three species composition of the mosquito fauna areas: water source reserve, the botanical in each stream. Laird (1988) described garden, and the Hapen Nature Preserve. species in the same community and their Annually air temperatures range from 4 possible interactions in various aquatic to 28 oC with an average of 20 oC. habitats used by mosquitoes. However, Average annual precipitation is 2900 mm flowing-stream studies were limited to of which 70% occurs in the summer highly polluted and stable drainage months. The rainy season occurs in ditches. winter, while typhoons bring most of the Efficient natural enemies of mosquito precipitation in summer. Sampling sites larvae in the laboratory are larvivorous were in the area of the Hapen Nature fish (Nelson and Keenan, 1992), Preserve, in which no tourists are damselfly naiads (Miura and Takahashi, allowed to enter. Four sections of the 1988), hemipteran bugs (Diplonychus spp.) creek in three types of habitats (ponded, (Venkatesan et al., 1986), hydrometrids blocked, and flowing streams) were (Palmer et al., 1998), notonedtid bugs chosen according to the sampling results (Wattal et al., 1996), copepods (Nasci et of the first month of this survey. The al., 1987), and freshwater prawns (Collins, characteristics of these four sampling 1998). Other natural enemies have also sites are described in Table 1. been recognized, such as tadpoles, dra- Larval Sampling. A 14-cm-diameter gonfly naiads, Culex fuscanus Widde- dipper was used to collect larvae. The mann, and C. halifaxii Theobald (Laird, dipper was held at an angle of 60o and 1988). 0.6 m from the stream bank. Eighty The Hapen Creek of northern Taiwan samples were taken in each section of the is highly protected from anthropogenic stream for a total of 320 dips per month influences but is frequently affected by (from September 1997 to August 1998), if flooding and drought. During the drought conditions permitted. Occasionally, no months we studied (October-December in samples were taken because of drought at 1997, and April, July, and August in a sampling site. Larvae were identified 1998), water flowed underground in some on site to species under 400X microscopes ௐαഇסᖪௐ˟˩αٿέ៉ 294 Table 1. Characteristics of the sampling sites in Hapen Creek, Ilan, Taiwan from September 1997 to August 1998 Habitat characteristics Ponded1 stream Blocked stream Flowing stream I Flowing stream II Water depth (cm) 7.5-99 8-37 9-40 6-35 Water velocity (cm/s) 0-9.5 0 0-22 0-21.4 Drought month 8 8 10-12, 4, 7-8 10-12, 4, 7-8 Overhead vegetation coverage Half open Dense vegetation Half open Dense vegetation Streambed Mud+sand Mud+sand Rocks Rocks 1Blocked stream was a side stream of a ponded stream; Flowing stream I was located 10~20 m uptream of Flowing stream II. and released back at the sampling points temperature, water velocity, water depth, except during for the last 2 months. or monthly rainfall). The population Pupae were brought back to the labora- density of mosquitoes is hard to quantify tory and reared to adults for species with the number of natural enemies in identification. At the same time, speci- the dip data, because most natural mens other than mosquitoes were collec- enemies do not necessarily come to the ted to check the species association with water surface to breathe. Therefore, only mosquito larvae. In addition, water tem- the absence or presence of dip data were perature, water velocity, and water depth analyzed by Spearman’s rank correlation were measured at each site. Monthly analysis to show the association of these rainfall data was obtained from the species to mosquito larvae. Spearman’s Fushan Weather Station. coefficient of rank correlation applies to Statistical Analysis. Species diver- data in the form of a rank (in our case, sity and species packing can estimate the the absence or presence in the dip data). stability of a habitat (Southwood, 1978). The rank-abundance plot or some other Results indices can describe species diversity, while associations of species or utilization Species Composition of Mosquito of resources can be used to estimate Fauna. In Hapen Creek, Anopheles species packing. Diversity indices were bengalensis Puri was the most abundant calculated for each habitat type using the species in terms of the total number Shannon-Wiener information index: collected and its wide distribution at all H’ = -∑(pi)log(pi); sampling sites (Fig. 1). Culex hayashii where pi is the proportion of the ith Yamada was the second most abundant species in the total sample (Shannon and was found in three out of four sites. 1948). Although An. lindesayi lindesayi Giles In this study, cluster analysis on the larvae were collected at all sampling basis of correlation coefficients was used sites, the size of the population was to group similar mosquito species, be- relatively small. Culex nigropunctatus cause no assumption is needed in this Edwards, Cx. halifaxii Theobald, Cx. analysis concerning the number of groups mimeticus Noe, Uranotaenia annandalei or the group structure (Johnson and Barraud, and An. gigas baileyi Edwards Wichern, 1988). In addition, correlation were found in small numbers at one or analysis was also used to check if the two sampling sites. Among the four number of larvae collected per dip was sampling sites, seven out of eight species associated with any of the physical existed in the blocked stream, however, environmental factors examined (water only two or three species were found in Mosquito Fauna of Hapen Creek 295 1000280 Anopheles bengalensis Culex hayashii 240800 Anopheles lindesayi Culex mimeticus 200600 Culex nigropunctatus Culex halifaxii 160400 Uranotaenia macferlanei Anopheles gigas baileyi 120200 80 Total no. of larvae collected 40 0 Ponded stream Blocked stream Flowing stream I Flowing stream II Habitat Fig. 1. Species composition of mosquito fauna in Hapen Creek Ilan, Taiwan. the ponded stream, flowing stream I, and lensis larvae increased from April to a flowing stream II (Fig. 2). A high number peak in June, and then decreased sharply of mosquito larvae was collected in ha- in July (Fig. 4). Another peak was found bitats of the ponded stream and blocked in November and December. For C. stream, but only a small population hayashii, two peaks were found in existed in the flowing stream I and October and July.